QuickMark

Project Overview

A Software is designed to automate the correction of bubble sheet exams for educational organizations. With a focus on efficiency, it significantly reduces the time and effort required for the correction process, empowers professors and teachers to allocate more time for more important tasks.

Technologies Used

• Core Language: Python
• Image Processing Libraries: OpenCV, NumPy
• Backend: Python Flask
• Frontend: HTML, CSS, Vue.js (under development)

Key Features

• Error Handling: Utilizes various techniques to address unclear sheets or non-correct shaded responses.
• Low Error Rate: Achieves an error rate of less than 0.5%.
• Speed: Rapidly corrects each sheet in about 200ms.
• Analysis: Provides insightful performance metrics for students after each exam.
• Simplicity: The correction process involves four straightforward steps: choose sheet model, upload model answer, upload sheets, download results, and analysis.

Challenges and Solutions

Encountered challenges, including low performance and difficulties in error handling, were solved through continuous improvement and algorithm refinement.

Project Structure

The project is organized into three main components:

1. Core Program: Manages the fundamental functionality of the correction process.
2. Backend: Implements server-side logic
3. Frontend: Currently under development

Testing and Validation

• The software is currently in beta and serves as the primary program for correcting bubble sheet exams at Borg Alarab Technological Unversity (Alex, Egypt).

Usage and Installation

• The software is not yet available for public usage as it is still in beta.
• For trying or more details, please contact me directly via email at [mostafamfelmasry@outlook.com].

Output samples

here is a final output file sample provide a full info about the corrected sheet(s).

columns explaination

• student ID : the id of the student
• Mark : sum of corrected question in the exam
• Page : i added this column for better exprience if the examiner wants to know where this sheet in the sheets file to manual review for example. imagine you want a sheet from file contains 500 sheet and you only have the student id, waste of time right?, for that reason i added this column
• Qx : the answer of the student in the question number x, there is four possible result (1 for true answer, 0 for false answer, M this appear if the student shade more than bubble and this treats a false answer, Z apear when the student shad 0 bubbles and this treats as false answer too)
  
  

This screenshot shows interactions with the backend with uploading sheets.pdf and get the result in result.zip through simple commands and clear feedback.

Future Plans

• Web app development is currently underway.